By Jason Charles Guck, Business Development, Delta Edge CI
There is one line on your electricity bill that you cannot do anything about this year, no matter how carefully you run the building. It was decided on a handful of hot afternoons last summer, and it will keep billing you every month until next spring.
That is the capacity charge, and it is the least understood cost in commercial energy. Most operators find it on the bill as an unexplained fixed amount and assume it is a tax. It is not. It is a measurement of your own building, taken at a specific moment, and the moment has already passed for the coming year.
Understanding that changes what you do about it. You stop trying to manage the charge and start managing the measurement.
What the charge actually is
Grid operators have to guarantee enough generating capacity to serve the worst hour of the year, not the average hour. Someone has to pay to keep that capacity available, and the cost is allocated to customers in proportion to how much each one contributed to the system’s peak.
Your share is expressed as a tag, and the tag is set by a measurement taken during the prior summer. Once set, it is fixed for the whole following year.
The mechanics differ by market, and the difference matters a great deal for what you can do about it.
New York: a single hour
In the NYISO market, your capacity tag is based on your consumption during the one hour when the whole New York system hit its annual peak. That peak hour is identified from July and August, on a non-holiday weekday. If your meter records interval data, the tag comes from your actual usage in that hour. If it does not, it is estimated from a load profile, which means you are assigned someone else’s shape whether or not it resembles yours.
The resulting tag then bills across a capability year that runs from May through April.
One hour. That is the entire measurement. A facility that happened to be running a full production shift during that hour pays as though it does so continuously.
PJM: five hours, averaged
In PJM territory the measurement is broader. Your peak load contribution is your average demand across the five highest system peak hours, drawn from June 1 through September 30, and the resulting obligation bills over a delivery year that runs June through May.
Five hours instead of one is more forgiving of a single bad afternoon, and it also means a single successful curtailment helps less. In New York you have one shot and it is worth a lot. In PJM you need to catch most of five.
Operators with sites in both markets are frequently managing them as though they work the same way. They do not, and a strategy tuned to one will underperform in the other.
Why this is worth an operator’s attention
Three things follow from the structure, and each one is actionable.
The charge is not proportional to your consumption. Two facilities using identical annual kilowatt-hours can carry very different capacity costs, because the measurement only looks at a few hours. A building that runs flat all year and a building that runs hard on hot afternoons are treated completely differently, and only one of them is being penalised.
Reducing it costs almost nothing in operational terms. You are not being asked to use less energy over the year. You are being asked to use less during a handful of hours that can be predicted a day ahead. For most commercial operations that is a scheduling question, not an investment question.
The saving persists. A tag reduced this summer bills lower for the following twelve months. That is unusual. Most conservation measures save while they run; this one is set once and then paid out over a year.
What to do between now and next June
The measurement window is closed for this year, which makes the next eight months the right time to prepare rather than the wrong time to worry.
- Find the charge on the bill and write down the tag. It appears under names like capacity, ICAP, installed capacity or peak load contribution, and it is usually stated in kilowatts. If you cannot find it, your supplier can tell you what tag they are billing against. You cannot manage a number you have never seen.
- Confirm the sites are interval-metered. If a site is billed on a load profile rather than its own interval data, its tag reflects an assumed shape. For a building that is unusually flat, or closed on summer afternoons, that assumption is probably costing money. Ask what it would take to change.
- Establish who receives peak-day notifications and what happens next. Suppliers and curtailment providers issue forecasts a day ahead. The information is generally available. What is usually missing is a named person who acts on it and a short written list of what gets turned down.
- Write the curtailment list now, in October, not in July. Decide in advance which loads can be shed for three or four hours without affecting production, service or comfort. Pre-cooling, deferring batch processes, staging non-essential equipment, adjusting setpoints within an agreed band. A list written under time pressure on a 95-degree afternoon is a list of arguments.
- Check whether curtailment can be paid for as well as avoided. The same reduction that lowers a capacity tag may also qualify for a demand response payment. The mechanics and the cases where it is genuinely worth it are covered in our post on demand response.
The honest limits
This does not work everywhere, and it is worth saying so plainly.
A facility whose load is genuinely non-discretionary during summer afternoons, a hospital, a data centre, a continuous process, has limited room to move and should not pretend otherwise. Peak forecasting is also imperfect: providers call more potential peak days than actually turn out to be peaks, so some curtailment effort is spent on days that did not count.
And the rules are set by tariff and by market rule, both of which change. Anything in this article should be confirmed against your own supplier’s current terms before you build a plan on it.
What does not change is the shape of the problem. A cost that is measured in a few hours and billed for a year rewards preparation more than it rewards effort. The organisations that do well on this are not the ones that work hardest in August. They are the ones that decided in October what they would do.
If you want your tags pulled, your interval data read against the actual system peaks, and a curtailment plan built before next summer, that is work we do at Delta Edge CI.